hcq/postgresql
1
0
Fork 0
mirror of https://git.postgresql.org/git/postgresql.git synced 2026-04-11 13:57:01 +08:00
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • 2b8758a389 Repair oversight in creation of "append relations": we should set up rel->tuples as well as rel->rows, since some estimation functions expect both to be valid in every baserel. Per report from Dave Dutcher. Tom Lane 2007-01-28 18:50:40 +00:00
  • 5681cde7b1 Make some small improvements in the accuracy of plpgsql's error location reports; inspired by the misleading CONTEXT lines shown in recent bug report from Stefan Kaltenbrunner. Also, allow statement-type names shown in these messages to be translated. Tom Lane 2007-01-28 17:58:13 +00:00
  • 17c8493c64 Fix path problem in MSVC bison wrapper. per Joachim Wieland. Andrew Dunstan 2007-01-28 16:29:37 +00:00
  • a534068e0e Add a new builtin type, "uuid". This implements a UUID type, similar to that defined in RFC 4122. This patch includes the basic implementation, plus regression tests. Documentation and perhaps some additional functionality will come later. Catversion bumped. Neil Conway 2007-01-28 16:16:54 +00:00
  • 8a114e0031 Fix up plpgsql's "simple expression" evaluation mechanism so that it behaves safely in the presence of subtransactions. To ensure that any ExprContext shutdown callbacks are called at the right times, we have to have a separate EState for each level of subtransaction. Per "TupleDesc reference leak" bug report from Stefan Kaltenbrunner. Tom Lane 2007-01-28 16:15:58 +00:00
  • a8e0b66061 Fix up plpgsql's "simple expression" evaluation mechanism so that it behaves safely in the presence of subtransactions. To ensure that any ExprContext shutdown callbacks are called at the right times, we have to have a separate EState for each level of subtransaction. Per "TupleDesc reference leak" bug report from Stefan Kaltenbrunner. Tom Lane 2007-01-28 16:15:49 +00:00
  • 76c7d2afee Drat, can't fit an additional argument into log_error. Is it worth an sprintf pushup to be sure we can report something useful for out-of-range exitstatus? Tom Lane 2007-01-28 07:29:32 +00:00
  • a9fa52504f Clean up broken usage of HAVE_DECL_SYS_SIGLIST and inconsistent/poorly formatted error messages. Tom Lane 2007-01-28 06:32:03 +00:00
  • 91ed399517 Use autoconf build-in sys_siglist macro AC_DECL_SYS_SIGLIST, rather than create our own. Bruce Momjian 2007-01-28 03:50:34 +00:00
  • 9bf559dee3 Add a delay at the start of the stats test, to let any prior stats activity quiesce. Possibly this will fix the large increase in non-reproducible stats test failures we've noted since turning on stats_row_level by default. Tom Lane 2007-01-28 03:02:31 +00:00
  • e96164dd71 Dept of second thoughts: the IQ of estimate_array_length() needs to be kept on par with that of scalararraysel(), else estimates that should track might not. Hence teach it about binary-compatible cases, too. Tom Lane 2007-01-28 02:53:42 +00:00
  • a053437d9e Dept of second thoughts: the IQ of estimate_array_length() needs to be kept on par with that of scalararraysel(), else estimates that should track might not. Hence teach it about binary-compatible cases, too. Tom Lane 2007-01-28 02:53:34 +00:00
  • c2c0b318d8 Add signal.h for sys_siglist reference. Bruce Momjian 2007-01-28 02:33:09 +00:00
  • ff3c2e4e46 Fix scalararraysel() to cope with binary-compatible cases, such as text[] versus varchar[]. This oversight probably explains Ryan Holmes' recent complaint --- he was getting a generic selectivity estimate instead of anything intelligent. Tom Lane 2007-01-28 01:37:45 +00:00
  • af18f6ad85 Fix scalararraysel() to cope with binary-compatible cases, such as text[] versus varchar[]. This oversight probably explains Ryan Holmes' recent complaint --- he was getting a generic selectivity estimate instead of anything intelligent. Tom Lane 2007-01-28 01:37:38 +00:00
  • 82480fc254 Use sys_siglist[] to print out signal names for signal exits, rather than just numbers. Bruce Momjian 2007-01-28 01:12:05 +00:00
  • 3ec7ae1b67 Modify SGML makefile to allow 'gmake draft' to build draft html ('draft' is now a target, no longer a modifier). Bruce Momjian 2007-01-27 22:44:32 +00:00
  • eea0902ead Correct an old logic error in btree page splitting: when considering a split exactly at the point where we need to insert a new item, the calculation used the wrong size for the "high key" of the new left page. This could lead to choosing an unworkable split, resulting in "PANIC: failed to add item to the left sibling" (or "right sibling") failure. Although this bug has been there a long time, it's very difficult to trigger a failure before 8.2, since there was generally a lot of free space on both sides of a chosen split. In 8.2, where the user-selected fill factor determines how much free space the code tries to leave, an unworkable split is much more likely. Report by Joe Conway, diagnosis and fix by Heikki Linnakangas. Tom Lane 2007-01-27 20:53:59 +00:00
  • 372c651312 Correct an old logic error in btree page splitting: when considering a split exactly at the point where we need to insert a new item, the calculation used the wrong size for the "high key" of the new left page. This could lead to choosing an unworkable split, resulting in "PANIC: failed to add item to the left sibling" (or "right sibling") failure. Although this bug has been there a long time, it's very difficult to trigger a failure before 8.2, since there was generally a lot of free space on both sides of a chosen split. In 8.2, where the user-selected fill factor determines how much free space the code tries to leave, an unworkable split is much more likely. Report by Joe Conway, diagnosis and fix by Heikki Linnakangas. Tom Lane 2007-01-27 20:53:52 +00:00
  • f109cb1285 Correct an old logic error in btree page splitting: when considering a split exactly at the point where we need to insert a new item, the calculation used the wrong size for the "high key" of the new left page. This could lead to choosing an unworkable split, resulting in "PANIC: failed to add item to the left sibling" (or "right sibling") failure. Although this bug has been there a long time, it's very difficult to trigger a failure before 8.2, since there was generally a lot of free space on both sides of a chosen split. In 8.2, where the user-selected fill factor determines how much free space the code tries to leave, an unworkable split is much more likely. Report by Joe Conway, diagnosis and fix by Heikki Linnakangas. Tom Lane 2007-01-27 20:53:46 +00:00
  • 5e6c06f040 Correct an old logic error in btree page splitting: when considering a split exactly at the point where we need to insert a new item, the calculation used the wrong size for the "high key" of the new left page. This could lead to choosing an unworkable split, resulting in "PANIC: failed to add item to the left sibling" (or "right sibling") failure. Although this bug has been there a long time, it's very difficult to trigger a failure before 8.2, since there was generally a lot of free space on both sides of a chosen split. In 8.2, where the user-selected fill factor determines how much free space the code tries to leave, an unworkable split is much more likely. Report by Joe Conway, diagnosis and fix by Heikki Linnakangas. Tom Lane 2007-01-27 20:53:41 +00:00
  • 223cd82c1d Correct an old logic error in btree page splitting: when considering a split exactly at the point where we need to insert a new item, the calculation used the wrong size for the "high key" of the new left page. This could lead to choosing an unworkable split, resulting in "PANIC: failed to add item to the left sibling" (or "right sibling") failure. Although this bug has been there a long time, it's very difficult to trigger a failure before 8.2, since there was generally a lot of free space on both sides of a chosen split. In 8.2, where the user-selected fill factor determines how much free space the code tries to leave, an unworkable split is much more likely. Report by Joe Conway, diagnosis and fix by Heikki Linnakangas. Tom Lane 2007-01-27 20:53:36 +00:00
  • 6cefacd7c8 Correct an old logic error in btree page splitting: when considering a split exactly at the point where we need to insert a new item, the calculation used the wrong size for the "high key" of the new left page. This could lead to choosing an unworkable split, resulting in "PANIC: failed to add item to the left sibling" (or "right sibling") failure. Although this bug has been there a long time, it's very difficult to trigger a failure before 8.2, since there was generally a lot of free space on both sides of a chosen split. In 8.2, where the user-selected fill factor determines how much free space the code tries to leave, an unworkable split is much more likely. Report by Joe Conway, diagnosis and fix by Heikki Linnakangas. Tom Lane 2007-01-27 20:53:30 +00:00
  • 722ad326f1 Back-port changes of Jan 16 and 17 to "revoke" pending fsync requests during DROP TABLE and DROP DATABASE. Should prevent unexpected "permission denied" failures on Windows, and is cleaner on other platforms too since we no longer have to take it on faith that ENOENT is okay during an fsync attempt. Tom Lane 2007-01-27 20:15:55 +00:00
  • cb476c1ec3 Back-port changes of Jan 16 and 17 to "revoke" pending fsync requests during DROP TABLE and DROP DATABASE. Should prevent unexpected "permission denied" failures on Windows, and is cleaner on other platforms too since we no longer have to take it on faith that ENOENT is okay during an fsync attempt. Tom Lane 2007-01-27 20:15:47 +00:00
  • 0fe1c36757 Fixed expected files, so they are in sync with tests again. Michael Meskes 2007-01-27 18:33:22 +00:00
  • ee57938c0b remove unnecessary and now inaccurate cast which I should have removed with other old code. Andrew Dunstan 2007-01-27 16:46:21 +00:00
  • 915abb346a Reactivate libxml memory management via palloc, now that I think I've classified the conditions under which this is safe to do (see source code comment). Peter Eisentraut 2007-01-27 14:50:51 +00:00
  • d3be7fae11 Add trailing zero byte in Unicode codepoint conversion. Peter Eisentraut 2007-01-27 11:48:31 +00:00
  • abbf860f72 Add: Bruce Momjian 2007-01-27 03:25:49 +00:00
  • ddf569e3ed Add: Bruce Momjian 2007-01-27 02:29:32 +00:00
  • f77b1f05f5 Add: Bruce Momjian 2007-01-27 02:28:16 +00:00
  • 175a242187 Allow args to spi_prepare to be standard type aliaes as well as those known in pg_type. Fixes bug #2917. Add some regression tests for these cases. Andrew Dunstan 2007-01-27 01:55:57 +00:00
  • 27552ce540 Update installation wording for an upgrade to state that dump/restore should not be done, per Peter. Bruce Momjian 2007-01-27 01:27:36 +00:00
  • adef25e5ea Control openjade draft-mode by variable DRAFT, rather than whether the version tag is 'devel'. Bruce Momjian 2007-01-26 23:51:39 +00:00
  • f0fc95612d Update wording of installation when upgrading, to more clearly distinguish major vs minor release upgrades. Bruce Momjian 2007-01-26 22:52:50 +00:00
  • 320abc3a95 Modify draft SGML instructions to use DRAFT=Y because recursion was causing html to be called twice --- no way to exit the makefile after the recursion returns. Bruce Momjian 2007-01-26 22:23:50 +00:00
  • 043fcd6616 On Windows, use pgwin32_waitforsinglesocket() instead of select() to wait for input in the stats collector. Our select() emulation is apparently buggy for UDP sockets :-(. This should resolve problems with stats collection (and hence autovacuum) failing under more than minimal load. Diagnosis and patch by Magnus Hagander. Tom Lane 2007-01-26 20:07:01 +00:00
  • 4355d214c2 On Windows, use pgwin32_waitforsinglesocket() instead of select() to wait for input in the stats collector. Our select() emulation is apparently buggy for UDP sockets :-(. This should resolve problems with stats collection (and hence autovacuum) failing under more than minimal load. Diagnosis and patch by Magnus Hagander. Tom Lane 2007-01-26 20:06:52 +00:00
  • 8ff2bccee3 Squelch some VC++ compiler warnings. Mark float literals with the "f" suffix, to distinguish them from doubles. Make some function declarations and definitions use the "const" qualifier for arguments consistently. Ignore warning 4102 ("unreferenced label"), because such warnings are always emitted by bison-generated code. Patch from Magnus Hagander. Neil Conway 2007-01-26 17:45:42 +00:00
  • 8924c56820 Update BSD/OS platform for 8.2. Bruce Momjian 2007-01-25 23:34:28 +00:00
  • 2e6d1e5f7a Add URL for shrinking tuple to just its headers: Bruce Momjian 2007-01-25 22:25:53 +00:00
  • 70268b50dd Update Win32 exception comment. Bruce Momjian 2007-01-25 21:50:49 +00:00
  • 74b6f73bc2 Add a setlocal command to the beginning of build.bat. This is required to deal with buildenv.bat properly, so that PATH (for example) doesn't expand infintly. Per report from Joachim Wieland, patch from Magnus. Neil Conway 2007-01-25 19:48:33 +00:00
  • a5a1506c96 Removed compiler warning due to unneeded unsigned declaration. Removed regression test that triggers those libc precision bugs on some archs. Michael Meskes 2007-01-25 16:45:25 +00:00
  • 167fa70a2e Update pg_dumpall -f option description. Bruce Momjian 2007-01-25 15:08:06 +00:00
  • c8bdd8ce88 Correction: temp_tablespaces was implemented by Albert Cervera Areny, with cleanup by Jaime Casanova. Bruce Momjian 2007-01-25 15:05:15 +00:00
  • 0626a7d1b5 Reverse out use of Py_RETURN_TRUE in plpython, only supported in Python >= 2.3. Bruce Momjian 2007-01-25 14:52:23 +00:00
  • 22bd156ff0 Various fixes in the logic of XML functions: Peter Eisentraut 2007-01-25 11:53:52 +00:00
  • 9597446d11 Done: Bruce Momjian 2007-01-25 04:38:59 +00:00
  • 148ea5cbea Add GUC temp_tablespaces to provide a default location for temporary objects. Bruce Momjian 2007-01-25 04:35:11 +00:00
  • ebe2830613 Properly detoast access to bytea field pg_trigger.tgargs. Old code might cause server crash. Bruce Momjian 2007-01-25 04:17:56 +00:00
  • 5af6b2abe9 Properly detoast access to bytea field pg_trigger.tgargs. Old code might cause server crash. Bruce Momjian 2007-01-25 04:17:46 +00:00
  • 251281767a Fix for plpython functions; return true/false for boolean, rather than 1/0. This helps when creating trigger functions that output SQL. Bruce Momjian 2007-01-25 04:08:51 +00:00
  • 6441288ec9 Add 'output file' option for pg_dumpall, especially useful for Win32, where output redirection of child processes (pg_dump) doesn't work. Bruce Momjian 2007-01-25 03:30:43 +00:00
  • 1b7d863f1d Remove developers list from TODO list now that we have URLs to reference discussions. Bruce Momjian 2007-01-25 02:50:12 +00:00
  • 7aa09c5df5 Done: Bruce Momjian 2007-01-25 02:48:06 +00:00
  • d37881fd90 Allow pg_dumpall to specify a database name rather than the default 'template1'. Bruce Momjian 2007-01-25 02:46:33 +00:00
  • 5ce94b28cc Add --tablespaces-only and --roles-only options to pg_dumpall. Bruce Momjian 2007-01-25 02:30:32 +00:00
  • ef65f6f7a4 Prevent WAL logging when COPY is done in the same transation that created it. Bruce Momjian 2007-01-25 02:17:26 +00:00
  • 693c85d954 When using MSVC, disable the building of ecpg if pthreads is not specified. Bruce Momjian 2007-01-24 19:24:28 +00:00
  • d56c800c40 Get pg_utf_mblen(), pg_utf2wchar_with_len(), and utf2ucs() all on the same page about the maximum UTF8 sequence length we support (4 bytes since 8.1, 3 before that). pg_utf2wchar_with_len never got updated to support 4-byte characters at all, and in any case had a buffer-overrun risk in that it could produce multiple pg_wchars from what mblen claims to be just one UTF8 character. The only reason we don't have a major security hole is that most callers allocate worst-case output buffers; the sole exception in released versions appears to be pre-8.2 iwchareq() (ie, ILIKE), which can be crashed due to zeroing out its return address --- but AFAICS that can't be exploited for anything more than a crash, due to inability to control what gets written there. Per report from James Russell and Michael Fuhr. Tom Lane 2007-01-24 17:12:47 +00:00
  • 4461eb17c6 Get pg_utf_mblen(), pg_utf2wchar_with_len(), and utf2ucs() all on the same page about the maximum UTF8 sequence length we support (4 bytes since 8.1, 3 before that). pg_utf2wchar_with_len never got updated to support 4-byte characters at all, and in any case had a buffer-overrun risk in that it could produce multiple pg_wchars from what mblen claims to be just one UTF8 character. The only reason we don't have a major security hole is that most callers allocate worst-case output buffers; the sole exception in released versions appears to be pre-8.2 iwchareq() (ie, ILIKE), which can be crashed due to zeroing out its return address --- but AFAICS that can't be exploited for anything more than a crash, due to inability to control what gets written there. Per report from James Russell and Michael Fuhr. Tom Lane 2007-01-24 17:12:41 +00:00
  • 6b8a7cfadc Get pg_utf_mblen(), pg_utf2wchar_with_len(), and utf2ucs() all on the same page about the maximum UTF8 sequence length we support (4 bytes since 8.1, 3 before that). pg_utf2wchar_with_len never got updated to support 4-byte characters at all, and in any case had a buffer-overrun risk in that it could produce multiple pg_wchars from what mblen claims to be just one UTF8 character. The only reason we don't have a major security hole is that most callers allocate worst-case output buffers; the sole exception in released versions appears to be pre-8.2 iwchareq() (ie, ILIKE), which can be crashed due to zeroing out its return address --- but AFAICS that can't be exploited for anything more than a crash, due to inability to control what gets written there. Per report from James Russell and Michael Fuhr. Tom Lane 2007-01-24 17:12:35 +00:00
  • 03d1281477 Get pg_utf_mblen(), pg_utf2wchar_with_len(), and utf2ucs() all on the same page about the maximum UTF8 sequence length we support (4 bytes since 8.1, 3 before that). pg_utf2wchar_with_len never got updated to support 4-byte characters at all, and in any case had a buffer-overrun risk in that it could produce multiple pg_wchars from what mblen claims to be just one UTF8 character. The only reason we don't have a major security hole is that most callers allocate worst-case output buffers; the sole exception in released versions appears to be pre-8.2 iwchareq() (ie, ILIKE), which can be crashed due to zeroing out its return address --- but AFAICS that can't be exploited for anything more than a crash, due to inability to control what gets written there. Per report from James Russell and Michael Fuhr. Tom Lane 2007-01-24 17:12:29 +00:00
  • 4e978c4e27 Get pg_utf_mblen(), pg_utf2wchar_with_len(), and utf2ucs() all on the same page about the maximum UTF8 sequence length we support (4 bytes since 8.1, 3 before that). pg_utf2wchar_with_len never got updated to support 4-byte characters at all, and in any case had a buffer-overrun risk in that it could produce multiple pg_wchars from what mblen claims to be just one UTF8 character. The only reason we don't have a major security hole is that most callers allocate worst-case output buffers; the sole exception in released versions appears to be pre-8.2 iwchareq() (ie, ILIKE), which can be crashed due to zeroing out its return address --- but AFAICS that can't be exploited for anything more than a crash, due to inability to control what gets written there. Per report from James Russell and Michael Fuhr. Tom Lane 2007-01-24 17:12:23 +00:00
  • 0887fa1117 Get pg_utf_mblen(), pg_utf2wchar_with_len(), and utf2ucs() all on the same page about the maximum UTF8 sequence length we support (4 bytes since 8.1, 3 before that). pg_utf2wchar_with_len never got updated to support 4-byte characters at all, and in any case had a buffer-overrun risk in that it could produce multiple pg_wchars from what mblen claims to be just one UTF8 character. The only reason we don't have a major security hole is that most callers allocate worst-case output buffers; the sole exception in released versions appears to be pre-8.2 iwchareq() (ie, ILIKE), which can be crashed due to zeroing out its return address --- but AFAICS that can't be exploited for anything more than a crash, due to inability to control what gets written there. Per report from James Russell and Michael Fuhr. Tom Lane 2007-01-24 17:12:17 +00:00
  • 234db5bae1 Relax an Assert() that has been found to be too strict in some situations involving unions of types having typmods. Variants of the failure are known to occur in 8.1 and up; not sure if it's possible in 8.0 and 7.4, but since the code exists that far back, I'll just patch 'em all. Per report from Brian Hurt. Tom Lane 2007-01-24 01:26:08 +00:00
  • 3e0b5417a5 Relax an Assert() that has been found to be too strict in some situations involving unions of types having typmods. Variants of the failure are known to occur in 8.1 and up; not sure if it's possible in 8.0 and 7.4, but since the code exists that far back, I'll just patch 'em all. Per report from Brian Hurt. Tom Lane 2007-01-24 01:26:02 +00:00
  • 212df03ac9 Relax an Assert() that has been found to be too strict in some situations involving unions of types having typmods. Variants of the failure are known to occur in 8.1 and up; not sure if it's possible in 8.0 and 7.4, but since the code exists that far back, I'll just patch 'em all. Per report from Brian Hurt. Tom Lane 2007-01-24 01:25:56 +00:00
  • 5edf3429ad Relax an Assert() that has been found to be too strict in some situations involving unions of types having typmods. Variants of the failure are known to occur in 8.1 and up; not sure if it's possible in 8.0 and 7.4, but since the code exists that far back, I'll just patch 'em all. Per report from Brian Hurt. Tom Lane 2007-01-24 01:25:51 +00:00
  • 07cf99ac6f Relax an Assert() that has been found to be too strict in some situations involving unions of types having typmods. Variants of the failure are known to occur in 8.1 and up; not sure if it's possible in 8.0 and 7.4, but since the code exists that far back, I'll just patch 'em all. Per report from Brian Hurt. Tom Lane 2007-01-24 01:25:47 +00:00
  • 4e8b5cd94b Simplify handling of XML error messages: Just use the string provided by libxml as the detail message. Peter Eisentraut 2007-01-23 23:39:16 +00:00
  • a56c5fb0f5 Update xindex.sgml to discuss operator families. Tom Lane 2007-01-23 20:45:28 +00:00
  • 379958128c Update pg_dump to support dumping operator families. Tom Lane 2007-01-23 17:54:50 +00:00
  • 867c133599 Add comment about exception lists in both winnt.h and ntstatus.h. Bruce Momjian 2007-01-23 16:21:17 +00:00
  • e25138f3e9 This patch is required for vcbuild to work after the changes to pg_proc.h (it's the same changes that's in gen_fmgrtab.sh in the unix build). Alvaro Herrera 2007-01-23 15:44:11 +00:00
  • a33cf1041f Add CREATE/ALTER/DROP OPERATOR FAMILY commands, also COMMENT ON OPERATOR FAMILY; and add FAMILY option to CREATE OPERATOR CLASS to allow adding a class to a pre-existing family. Per previous discussion. Man, what a tedious lot of cutting and pasting ... Tom Lane 2007-01-23 05:07:18 +00:00
  • 8502b68513 Remove newline from error message because URL is gone. Bruce Momjian 2007-01-23 03:31:33 +00:00
  • 882b9948d7 Back out use of FormatMessage(), does error values, not exception values. Point to /include/ntstatus.h for an exception list, rather than a URL. Bruce Momjian 2007-01-23 03:28:49 +00:00
  • 610f60a092 Print meaningfull error text for abonormal process exit on Win32, rather than hex codes, using FormatMessage(). Bruce Momjian 2007-01-23 01:45:11 +00:00
  • b97b86649a Update my email address in FAQ. Bruce Momjian 2007-01-22 23:06:37 +00:00
  • 4f06c688c7 Put back planner's ability to cache the results of mergejoinscansel(), which I had removed in the first cut of the EquivalenceClass rewrite to simplify that patch a little. But it's still important --- in a four-way join problem mergejoinscansel() was eating about 40% of the planning time according to gprof. Also, improve the EquivalenceClass code to re-use join RestrictInfos rather than generating fresh ones for each join considered. This saves some memory space but more importantly improves the effectiveness of caching planning info in RestrictInfos. Tom Lane 2007-01-22 20:00:40 +00:00
  • 45e0736938 Use errhint() for WIN32 SIGTERM message, where possible. Bruce Momjian 2007-01-22 19:38:05 +00:00
  • d26a5f1fea Uppercase hex value. Bruce Momjian 2007-01-22 18:32:57 +00:00
  • 208ae0c290 When system() fails in Win32, report it as an exception, print the exception value in hex, and give a URL where the value can be looked-up. Bruce Momjian 2007-01-22 18:31:51 +00:00
  • eeae929dbe Update documenation instructions. on how to delete template1. Bruce Momjian 2007-01-22 02:47:56 +00:00
  • 686f15e39e Adjust pgbench so it won't spit up on non-select queries returning tuples, which is entirely possible with custom scripts (consider RETURNING, EXPLAIN, etc). Tom Lane 2007-01-22 02:17:30 +00:00
  • 5a7471c307 Add COST and ROWS options to CREATE/ALTER FUNCTION, plus underlying pg_proc columns procost and prorows, to allow simple user adjustment of the estimated cost of a function call, as well as control of the estimated number of rows returned by a set-returning function. We might eventually wish to extend this to allow function-specific estimation routines, but there seems to be consensus that we should try a simple constant estimate first. In particular this provides a relatively simple way to control the order in which different WHERE clauses are applied in a plan node, which is a Good Thing in view of the fact that the recent EquivalenceClass planner rewrite made that much less predictable than before. Tom Lane 2007-01-22 01:35:23 +00:00
  • a85e9c61e5 Remove some unused variables Peter Eisentraut 2007-01-21 09:23:29 +00:00
  • 962e825cd7 Put back libpq link, but make sure that libpq's in-tree directory is mentioned before the system directories. Peter Eisentraut 2007-01-21 09:19:06 +00:00
  • 066926dfbb Refactor some lsyscache routines to eliminate duplicate code and save a couple of syscache lookups in make_pathkey_from_sortinfo(). Tom Lane 2007-01-21 00:57:15 +00:00
  • 98218e9ef6 Add bitmap index URL: Bruce Momjian 2007-01-20 23:18:38 +00:00
  • fcf4b146c6 Simplify pg_am representation of ordering-capable access methods: provide just a boolean 'amcanorder', instead of fields that specify the sort operator strategy numbers. We have decided to require ordering-capable AMs to use btree-compatible strategy numbers, so the old fields are overkill (and indeed misleading about what's allowed). Tom Lane 2007-01-20 23:13:01 +00:00
  • c82cc604f5 Add URL for dead space map: Bruce Momjian 2007-01-20 23:08:20 +00:00
  • 726f0807ad Update docs for recent change to setseed() return value. Neil Conway 2007-01-20 21:59:34 +00:00
  • aef0f53b62 Make setseed() return void, rather than an int4 without any use. Per pgsql-patches discussion of September 20, 2006. Bump the catversion. Neil Conway 2007-01-20 21:47:10 +00:00
  • 9352ed9a9c Add documentation of memory and time units to postgresql.conf. Backpatch to 8.2.X for new initdbs. Bruce Momjian 2007-01-20 21:42:06 +00:00
  • ebef17c7a2 Add documentation of memory and time units to postgresql.conf. Backpatch to 8.2.X for new initdbs. Bruce Momjian 2007-01-20 21:42:03 +00:00
  • 67f06327f9 Don't use 'int' where you mean 'bool'. Tom Lane 2007-01-20 21:40:25 +00:00